Energy markets are inherently volatile, influenced by technological advancements, policy changes, global events, and consumer trends. Accurate quantification of the "price of energy" is crucial for evaluating the effectiveness of energy and monetary policies. Existing tools lack comprehensive coverage of the entire US energy landscape, often focusing on specific sectors or commodities. This research addresses this gap by developing EPIC, a novel predictive framework that calculates the average price of energy across all energy sources and end-use sectors in the United States. This index serves as a benchmark for analyzing the impact of various policies on energy prices and provides a holistic view of the US energy market, unlike existing indices that focus on specific sectors or commodities such as oil and gas, or the performance of energy companies. The study utilizes a rolling horizon methodology to predict future energy demands, enabling the assessment and optimization of energy policies.

The existing literature offers various approaches to modeling and forecasting energy prices, but these are often limited in scope. Some studies focus on specific energy sectors (electricity, natural gas, crude oil, petroleum products), using methods like double seasonal exponential smoothing, cointegration, and vector error correction models. Forecasting horizons are generally short. Existing energy indices primarily rely on market capitalization or production weights (e.g., S&P 500 Energy Index, MSCI US IMI Energy 25/50 Index, S&P GSCI Energy), failing to capture the actual prices of all energy products and their demand across the entire energy landscape. This research distinguishes itself by providing a holistic approach that integrates both supply and demand factors across all major energy sectors, offering a longer-term forecasting capability and a more accurate representation of the average energy price to consumers.

The EPIC framework calculates the average price of energy using two key factors: total demand of energy products directed to end-use sectors (residential, commercial, industrial, transportation) and the corresponding price of each product. Data on energy consumption and prices are sourced from the US Energy Information Administration (EIA) and other publicly available datasets. The complex US energy landscape is carefully analyzed to avoid double-counting, focusing on energy products reaching the end-use sectors directly. The electric power sector is treated as an intermediate sector, with its output (electricity) considered as a product consumed by the four end-use sectors. A rolling horizon predictive methodology is implemented to estimate future energy demands. This approach uses data from the past three time periods to predict the weights of energy demand for the current and future periods. Four different approaches for minimizing the sum of squared errors were tested across four different lookback periods. Approach 2 (weight-based) with a 36-month lookback period showed the lowest error and was selected. The model minimizes the squared difference between the real value of a product's weight in the past and its predicted value for the month of interest. The accuracy of this approach is validated by comparing predicted weights to actual weights over a 174-month period. Two policy case studies are presented: **Policy Case Study 1 (Crude Oil Tax):** This study parametrically investigates the effects of a crude oil tax (ranging from $2.5 to $25 per barrel) on EPIC and government revenue, both historically (2003-2020) and prospectively (2020-2024). The impact on household energy expenses is also calculated. In the model, crude oil demand is considered inelastic in the short-run. **Policy Case Study 2 (Renewable Energy Subsidies):** This analysis assesses the impact of various targets for renewable energy feedstocks (nuclear, hydroelectric, biomass, geothermal, solar, and wind) within the electric power sector. The analysis is done with a wide range of tax credits (0-$9/MMBtu) and explores their effect on EPIC and government budget requirements, both historically and prospectively. Levelized costs of electricity production are incorporated from Lazard's Levelized Cost of Energy Analysis report and EIA Annual Energy Outlook.

The rolling horizon methodology demonstrated excellent predictive ability with very low prediction errors (below 2.4% even for the fourth year of prediction). EPIC, as calculated, represents a comprehensive average price of energy in the US, considering all energy sources and end-use sectors. **Policy Case Study 1 (Crude Oil Tax):** A $10.25 per barrel crude oil tax increased EPIC by approximately 5.6% historically and 4.2% in the projected period (2020-2024), generating substantial revenue. The average annual revenue from a $10.25 per barrel tax was estimated at $70.962 billion. A $2.5 per barrel increase in crude oil tax led to an 1.35% rise in household energy expenses in 2015. **Policy Case Study 2 (Renewable Energy Subsidies):** The analysis reveals that increased weights of hydroelectric, wind, solar, and geothermal energy in the power sector generally decrease EPIC, even without tax credits. Nuclear and biomass required tax credits to lower EPIC. The highest impact in reducing EPIC comes from subsidies for nuclear, hydroelectric, and wind power. The budget requirements for the subsidies vary greatly depending on the chosen target and tax credit level, with nuclear requiring the most significant investment.

The EPIC framework provides a valuable tool for policymakers to assess the quantitative effects of various energy policies. The excellent predictive capability allows for an accurate estimation of future energy prices and policy impacts. The case studies demonstrate the framework's ability to analyze the trade-offs between different energy sources and policy instruments (taxes versus subsidies). The findings highlight the potential for significant revenue generation through crude oil taxation and the potential cost-savings for consumers through increased reliance on renewable energy sources like hydroelectric and wind, which demonstrates cost-effectiveness even without subsidies. The findings could inform policy decisions aimed at balancing environmental goals with economic considerations.

The Energy Price Index (EPIC) offers a novel and comprehensive framework for assessing the average price of energy in the US and its response to policy changes. Its predictive capabilities and holistic approach improve upon existing methodologies. Future research could incorporate more sophisticated modeling of financial shocks, technological advancements, and detailed simulations of the interactions between multiple energy sources within a single framework. Incorporating AI methods could enhance forecasting accuracy. The ultimate goal is to utilize EPIC for optimizing federal renewable energy policy to mitigate climate change while keeping energy prices affordable.

The model assumes inelastic crude oil demand in the short run, which might not always hold true. The levelized costs of energy feedstocks used in the renewable energy case study rely on historical data and may not perfectly capture future cost fluctuations due to technological progress or policy incentives. The analysis does not account for all potential indirect effects of the proposed policies on other sectors of the economy or the social and environmental consequences beyond financial parameters.